Baseline composition and microbial quality assessment of raw milk from endemic small ruminants and Maghrebi camels in the oasis area of Tunisia

Characteristics and quality aspects of milk from native ovine queue ne de l’Ouest (QFO) and the local goat population were investigated and compared with those of the local Maghrebi camel. A total of 378 individual milk samples were collected from lactating animals reared in the continental oasis region of Tunisia. Samples were analyzed for physical parameters (pH, density, and acidity), chemical composition (dry matter, fat, protein, lactose, casein, ash, and casein-protein ratio), mineral concentrations (Ca, P, Na, and K) and microbiological features (total mesophilic aerobic bacteria (TMAB), total coliform count (TCC), lactic acid bacteria (LAB), sulphite-reducing Clostridium (CSR), yeast and molds (Y/M), Staphylococcus aureus (S. aureus), Escherichia coli (E. coli), and Salmonella) according to standard methods. The results obtained for milk characteristics revealed noticeable disparities between the three species. The mean values of pH, density, and acidity in milk collected from sheep appeared higher than those in milk collected from goat species. Compared to the camel populations, sheep species produced milk with similar pH but higher density and acidity. Between camel and goat species, pH and acidity were higher in Negga, while the density was similar. For milk composition, the results showed a remarkable variation among all studied species and an obvious superiority of the ovine species over the caprine and camel populations in all the chemical contents being studied, except for the casein-protein ratio, which is in favor of goat species. The milk of QFO sheep, the richest in casein and protein, was expressed with signicantly higher levels of calcium and phosphorus than goat and camel milk. Compared to small ruminants, milk from camels is the richest in Na and K. Additionally, more Ca is present in milk from camels than goats. Goat milk, the poorest type of milk in Ca and Na, contains on average more P than camel milk and more K than sheep's milk. The poor bacteriological quality was that of camel milk for all microbial counts. The microbial quality of goat milk was higher than that of ewe milk based on TMAB, TCC, and E. coli counts, while ovine milk was of better quality, referring to LAB, Y/M, and S. aureus values. No signicant differences were found for Staphylococcus aureus and Escherichia coli between the examined species. The obtained results highlighted the complete absence of the two dangerous pathogens Salmonella and CSR in all investigated milk samples. The microbiological examination evidenced that the milk of small ruminant species complies with standard criteria required by Tunisian legislation on the hygiene of milk and dairy products. Regarding camel milk, the microbial analysis revealed poor quality that exceeds standard criteria.

milking intended for local consumption (Vacca et al. 2009) or transformation for local products occurs (Leben, D'hen, Rayeb…) (Gaddour et al. 2014). The camel species, which is the most suitable animal anatomically and physiologically adapted to a harsh and painful drought environment, is entirely composed of the Maghrebi population. Animals are raised mainly within two management systems in the country's south, a traditional pastoral system and a semi-intensive system created in response to the decline of pasture (Fguiri et al. 2018). Camels are traditionally used for meat production (Chamekh et al. 2020), and substantial milking is intended for local consumption (Ayadi et al. 2009). Over recent decades, in oasis regions, a camel milk sector for marketing and human consumption has emerged throughout Tunisia.
It is well documented that the milk of all mammals contains the same principal components, but their content varies widely between species (Roy et al. 2020). Even within the same species, the milk composition may vary considerably, given the in uence of genetic factors (not only at the species level but also at the breed level), physiological factors (e.g., lactation stage, milking interval, parity), nutritional factors (e.g., feed energy value and composition), and environmental conditions (e.g., location, season) (Claeys et al. 2014).
Except for the dairy Sicilo-Sarde sheep breed, little information is available on the characteristics of milk from mammalian species in Tunisia. Such studies on physicochemical composition and bacteriological quality are extremely  Kondyli et al. 2012) and the marketing of products (Park et al. 2007). Investigation of milk characteristics is necessary to provide information essential to the development of e cient management plans for sustainable use and the improvement of these underexploited local genetic resources. Thus, this study aimed to assess and compare the physical characteristics, chemical composition, mineral concentrations, and microbiological quality of ewe, goat, and Negga milk in Tunisian oasis areas.

Animals and samples collection
Species of goats, sheep, and one-humped camel from private herds reared in the continental oases region of southwest Tunisia were used in this study, which was conducted in 2018 and 2019. These included milk samples from the QFO sheep breed (N= 100), local goat population (N= 229), and Maghrebi breed (N= 49) for camel species. Before sample collection, basic data on the animals (health status, age, calving season, lactation stage, parity number, etc.) were collected from the breeders. Small ruminant animals were selected for this study according to uniform body condition, suitable health status, equable age (4 years), parity (multiparous), lactation number (3rd lactation), and at the midlactation stage. Healthy camels with an average age of 9.5 ± 3.6 years, variable parities, and at the mid-lactation stage (the second and third months of lactation) were used in this study.
The QFO sheep and local goats were raised under a traditional feeding system where herds grazed on natural pasture in the oases' vicinity during the daytime (6-7 h/day) and were housed the rest of the day and overnight. Once back from grazing, animals were given 300-400 gd - 14.20%; Net energy content: 1320 Kcal. kg -1 DM). For both species, females were kept together with their lambs/kids while in con nement during the day and in the evening, except on days when milk samples were taken.
The selected camels were maintained under traditional system management where herds grazed on natural desert pastures were mainly characterized by halophilic species around chotts and nonhalophilic plants such as Haloxylon salicornicum, Anabasis articulata, Atriplex mollis, Atriplex halimus, Retama Raetam, Haloxylan schmittianum, Panicum turgidum, Traganum nudatum, Calligonum comosum, Calligonum azel, Aristida Pungens, Limoniastrum guyonianum, Stipagrostis pungen, Zygophyllum album, Sueda fructosa, Tamarix aphylla, Tamarix articulata, Tamarix gallica, Rhus tripartitum, Ephedra alata, and other endemic plants. In times of scarcity, animals were given supplementation based on barley, wheat bran throughout the year, and dates during the period from October to June when wasted dates were available. All feedstuffs were distributed to all animals without respect to their physiological stage.
A total of 378 random samples of fresh milk were collected in the morning by direct milking from complete milking. A duplicate individual sample of 300 ml from healthy animals was collected in two sterile tubes and kept in an ice container during sampling and transportation to the laboratory. The rst tube was brought to the animal production laboratory (CRRAO) for physicochemical and mineral analysis. The second tube was transported directly to the regional public health laboratory of Tozeur for microbiological analysis.

Physical and chemical analyses
Physical parameters (pH, density, and dornic acid) were determined during the same sampling day. The pH was measured at 20°C using a Consort C933 pH meter. To determine raw milk density, a Gerber thermolacto-densimeter was used, and measurements were made at 20°C. Dornic acidity was determined using the titrimetric method outlined in AOAC (2000). The milk samples were analyzed using o cial AOAC International analytical methods for lactose (AOAC, 2005) and ash (AOAC, 2012). Following the IDF Standard Methods, fat (IDF, 2009), dry matter (IDF, 2010), and total protein (IDF, 2014) were determined. The casein content was determined by the difference between the total nitrogen and the non-casein nitrogen determined by the Kjeldahl method (IDF, 2004). The casein/total nitrogenous matter ratio was determined to assess the cheese value of the studied milk.

Mineral analysis
Milk samples were analyzed for macrominerals, including calcium, sodium, potassium, and phosphorus. Calcium was measured according to IDF (2007) using an atomic absorption spectrophotometer (Analytikjena: nova 400). The determination of sodium and potassium was performed by Jenway ame emission spectroscopy in accordance with AOAC (1984). The colorimetric method involving the PhosphoVanado Molybdate complex (GB, 2010) was applied to quantify the phosphorus present in the milk sample.
After properly mixing the raw milk samples, 1 ml was taken, and dilutions with 9 ml of peptone water were prepared for microbiological analyses. provided as a colony-forming unit per ml (CFU/ml). The bacterial counts were log 10 -transformed to normalize the distributions before performing statistical analysis.

Statistical analysis
Statistical analysis was performed by comparing the averages of different parameters between species being studied. The signi cant differences between means were determined by one-way analysis of variance (ANOVA) using the general linear model (GLM) procedure of SAS software (2004). The difference between the three species was determined by comparing the least-square means with adjusting the P values for multiple comparisons using the Tukey-Kramer test.

Results
The physical characteristics of the milk samples summarized in Table (1) revealed a signi cant difference between the three species (P<0.001). Compared with goat's and camel's milk, sheep's milk had a higher density and dornic acidity (P<0.001). The results showed that the highest pH was obtained equally (P > 0.05) for ovine and camel species, followed by goat species, which presented the lowest pH (P<0.001). The highest dornic acidity was observed in milk from sheep, followed by camel and then goat species (P < 0.001). The ovine QFO breed produced the densest milk (P < 0.001), and the least dense milk was that of the goats and Maghrebi camel, which did not differ signi cantly (P > 0.05). The chemical composition of milk differed signi cantly (P < 0.001) between the species under study (Table 2). Sheep's milk had the highest (P < 0.001) dry matter, fat, protein, lactose, casein, and ash contents. The local goat population occupies the second ranking for all study components. Finally, the camel population recorded the lowest values for chemical constituents under study.
Although ovine species were superior in terms of casein and nitrogen content, the casein-protein ratio was higher among caprine species (0.80), followed by sheep breed (0.77) and last camel population (0.74) (P < 0.001). As depicted in Table 3, species exhibited signi cant differences (P < 0.001) in calcium, phosphorus, sodium, and potassium concentrations. Regarding mineral concentrations, the results showed that milk produced by ewes had the highest contents of calcium and phosphorus (P < 0.001). Compared to the goat species in terms of calcium and phosphorus, camel milk was richer in calcium (P < 0.001), while goat's milk contained more phosphorus (P < 0.001). In contrast, for the other major minerals, Negga milk was found to be richer in sodium and potassium (P < 0.001) than small ruminants. Ewes produced milk with more sodium concentration and less potassium than goats (P < 0.001). The bacteriological quality of different types of milk is given in Table 4. According to the results, the microbial quality of goat's milk was higher (P < 0.001) than that of sheep and camel milk based on TMAB, TCC, and E. coli counts. Ovine milk was of better quality, referring to LAB, Y/M, and S. aureus values (P < 0.001). The poor bacteriological quality was that of camel milk for all microbial counts.
The total mesophilic aerobic bacteria (TMAB) and Escherichia coli counts in the Maghrebi camel were signi cantly higher (P <0.001) than those in the QFO sheep and local goats. Based on the total coliform (TCC) and S. aureus counts, camels still recorded higher levels of contamination (P < 0.001) than those of sheep and goat species, which did not differ signi cantly (P > 0.05).
Lactic acid bacteria (LAB) and molds and yeasts (Y/M) values enumerated in this study for each species were higher in camel milk than in goats and QFO sheep at a highly signi cant level (P < 0.001).
The three species analyzed were signi cantly different (P < 0.001) from each other for the average molds and yeasts (Y/M) and lactic acid bacteria (LAB) values. Levels of contamination were higher in camel milk than in goat and QFO sheep. In all investigated samples, pathogens, sulphite-reducing Clostridium (CSR), and Salmonella were absent throughout.   (Farah, 1993). The results inferred that the lowest concentration in whey protein was that of goat milk, and the highest concentration was that of camel milk (Hilali et al. 2011). The increase in whey protein has technological implications, such as a weaker texture of curd and lower cheese yield (Barlowska et al. 2020). Proteins are a determinant factor affecting the quality of dairy products, as the reduction in proteins and casein contents results in poor dairy technology-making properties (Hilali et al. 2011). In another way, a lower casein-to-whey-protein ratio (i.e., a higher proportion of whey proteins) has been shown to be more desirable for faster digestion of the milk proteins in infant formula than a casein-dominant protein composition (Roy et al. 2020), which is the case of camel milk, recently declared very similar to human milk in terms of qualitative whey protein pro le (El-Hatmi et al. 2015) and as a most suitable substitute for cow milk when considering the preparation of infant formulas (Mudgil et al. 2022).

Physical characteristics
On the whole samples, the mean value of all physical characteristics (pH, density, and Dornic acidity) in milk collected from sheep breeds appeared higher than in milk collected from goat species. Compared to the camel populations, sheep species produced milk with a similar pH but a higher density and acidity content. Between camel and goat species, pH and acidity were higher in Negga, whereas the density was identical in both species. The physical properties of milk are widely reported to be associated with its composition and animal species (Park et al., 2007;Hilali et al. 2011). The fat content associated with total solids in milk has a determining in uence on its density, as has already been explained by other authors. (Park et al. 2007).

Mineral concentration
The variations in mineral concentrations among animal species are indicated in Table 3. The current results revealed considerable differences in the mineral concentrations of milk from different species. The levels of Ca and P are higher in sheep than in goat and camel milk. Compared to small ruminants, milk from the camel is the richest in Na and K. Additionally, more Ca is present in milk from camels than goats. Goat milk, the poorest type of milk in Ca and Na, contains on average more P than camel milk and more K than sheep's milk.
When comparing sheep breed versus goat milk for macroelements, the current study found lower levels

Microbiological features
The microbial quality of goat milk was higher than that of ewes and Negga's milk based on TMAB, TCC, and E. coli counts. Ovine milk is of better quality, referring to LAB, Y/M, and S. aureus values. The poor bacteriological quality was that of camel milk for all microbial counts.
S. aureus and E. coli were prevalent in all milk types with different levels of contamination. Two primary sources cause Staphylococcus in milk: the rst is the lack of adequate hygiene measures and inappropriate handling during milking (Fatima et al. 2013), whereas the second is mastitis, which affects animals (Benmeziane-Derradji, 2021).
In the current study, the selected animals were healthy and milked respecting hygienic practices; thus, Staphylococcus prevalence in the milk samples may be linked to subclinical mastitis occurrence (Alebie et al. 2021).
No signi cant differences in the numbers of total coliform populations (TCCs) were noted between the milk of the studied species, except Maghrebi animals, which generate milk with the highest count. Higher TMAB, TCC, LAB, Y/M, S. aureus, and E. coli counts in ewe milk were obtained by Fatima et al. (2013). In goat milk, Tabet  High total bacterial counts in raw milk mainly re ect the poor hygienic condition under which the milk was handled, storage temperature and time elapsed since milking, and the poor health of milking animals (Adugna et al. 2013). With the current study, the main source of contamination could be attributed to the contamination of the camel udder by the hands of unhygienic milkers or unhygienic milking procedures. Microorganisms can be transferred from the environment, i.e., feces, bedding, and soil, from contaminated hands, clothing, and mouth of milk handling personnel (Alebie et al. The results from the current study showed that the levels of microbial contamination of raw small ruminant milk in the oasis regions of Tunisia were satisfactory. Microbiological analysis meets the standard criteria required by Tunisian legislation on the hygiene of milk and dairy products (NT 14.141 (2004)). In contrast, the results showed that the levels of microbial contamination of raw camel milk in the study area were unsatisfactory and could not comply with the standard requirements of Tunisian legislation. Camel milk is commonly produced, conserved, and transported under unhygienic conditions. The bacteriological quality of raw milk should therefore be a major concern for farmers, processors, and the general public because bacteria in milk can degrade milk components, decrease shelf life, and cause illnesses in human beings (Adugna et al. 2013).

Conclusion
The current results contribute to the characterization of local camel, goat, and sheep breeds raised in Tunisian oases regions regarding physicochemical composition, mineral content, and bacteriological properties of milk. The results revealed notable differences between the three milk types and the particular characteristics that differentiate each species. Indeed, the QFO sheep breed generates the best milk in terms of density, acidity, dry matter, fat, protein, casein, ash, Ca, P, LAB, Y/M, and S. aureus count. The Na and K concentrations distinguished the Maghrebi camel, where the local goat population produces milk with relevant casein-protein ratios, TMAB, TCC, and E. coli counts. The analysis of small ruminants and camel milk reveals good physicochemical characteristics and appreciable mineral pro les compared to standards encountered in the scienti c literature. The bacteriological quality was satisfactory and complied with the standards criteria required by Tunisian legislation for small ruminant milk. For camel milk, the microbial analysis revealed poor quality that exceeds the standards, requiring strict hygienic control along the value chain.
With such interesting properties, the milk of local animal genetic resources in oasis areas is ideal for manufactured dairy products, which creates excellent opportunities for small breeders to increase their incomes, value autochthonous breeds and preserve their biodiversity.

Declarations Funding
This work was supported nancially by the Regional Center for Research in Oasis Agriculture (CRRAO). The funding body had no role in the design of the study and collection, analysis, and interpretation of data or in writing the manuscript.

Con icts of interest/Competing interests
The authors declare that they have no competing interests.

Availability of data and materials
The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.

Code availability
Not applicable Authors' contributions Zahran Khaldi was responsible for the study design, milk sample collection, physicochemical and mineral analysis, contributed to the microbiological analysis, performed statistical analysis and interpretation of data, drafted the manuscript, and was a major contributor in writing the manuscript. Zahran Khaldi is the corresponding author.
Mounir Nafti participated in the study design, and the milk sample collection, physicochemical, and mineral analysis contributed to the microbiological analysis, analysis, and interpretation of data and was a contributor in writing the manuscript.
Mohamed Tabarek Jilani took part in the design of the study, performed the microbiological analysis of milk samples, and participated in the data interpretation.
Zahran Khaldi written the rst draft of the manuscript, and all authors commented on previous versions of the manuscript. All authors read and approved the nal manuscript.